A heterologous remote sensing image registration method and device of a conditional generative adversarial network

Abstract

The invention discloses a conditional generative adversarial network-based heterologous remote sensing image registration method and device based on image segmentation, and the method comprises the steps: obtaining a training sample according to two groups of heterologous remote sensing image data; Performing image segmentation on the input image to obtain a segmented image; Constructing a conditional expression generative adversarial network based on image segmentation; Constructing a generative adversarial network of the image blocks; Training the two generative adversarial networks; And performing homologous image registration on the generated image and the to-be-registered image. According to the method, the spatial information consistency of the generative adversarial network generated images can be constrained, an image conversion solution with full-size spatial information consistency is provided, a heterologous image registration problem is simplified into a homologous image registration problem, and a unified heterologous remote sensing image registration framework is constructed.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a heterogeneous remote sensing image registration method and device based on conditional generative adversarial networks for image segmentation. Background technique [0002] Heterogeneous remote sensing image registration is a technology to register the geometric information of data acquired by different sensors in the same area. It is a key technology for the automatic processing of massive remote sensing data. Its calculation accuracy will directly affect the fusion and mosaic of heterogeneous remote sensing images. The precision of the operation to be applied subsequently. However, due to differences in imaging mechanism, shooting time, satellite attitude, and sensor parameters, there are often large differences in size, rotation, and translation between the data acquired by sensors of different nature. At the same time, there are often non-linear difference...

AI Technical Summary

Problems solved by technology

At present, the heterogeneous remote sensing registration method based on machine learning needs to use image conversion based on generative confrontation network, but there is no strict constraint on image conversion to match the spatial information consistency of image blocks, so it is easy to cause mismatch of image blocks

Images